In this week’s GeriPal podcast we interview Randy Curtis, Professor of Medicine and Pulmonary Critical Care and Director of the Palliative Care Center of Excellence at the University of Washington.
The issue is – how do we get more patients with chronic and serious illness to engage in goals of care conversations with their doctors? One way is to train doctors in better communication techniques. We’ve talked with folks – most recently Wendy Anderson – about Vital Talk and other such methods that approach the issue from that angle.
While acknowledging the critical importance of improving the quality of clinician communication, Randy’s focus in the present study is priming patients and clinicians to engage in these conversations. So how do you prime them? Listen to the podcast to find out! (bonus attempt to play Ed Sheeran)
Brief hint – patients are surveyed in advance to ascertain readiness and barriers to goals of care conversations. Then, and this is brilliant, the clinician is given feedback about the specific barriers, and advice drawn from VitalTalk about how to approach these barriers.
Eric: Welcome to the GeriPal podcast. This is Eric Widera.
Alex: This is Alex Smith.
Eric: And Alex, who is our special guest today.
Alex: Today we have a guest from the University of Washington. We have Randy Curtis, an esteemed and established palliative care researcher. He’s Professor of Medicine and Pulmonary Critical Care up there and Director of the Palliative Care Center of Excellence. Welcome to the GeriPal podcast, Randy.
Randy: Great. I’m excited to be here.
Eric: So before we get into our topic of the day, we always ask our guests if you have a song for Alex to sing.
Randy: I do. I consulted my 14 year old daughter and we picked Supermarket Flowers by Ed Sheeran.
Eric: And I understand that you saw Ed Sheeran perform on your sabbatical in Paris recently. Is that right?
Randy: Yeah, my daughter and I went to see him live in Paris. It was pretty awesome.
Alex: Is he the one who likes to rap occasionally in his concerts?
Randy: He does. He raps and he also does the entire concert as a one person show.
Alex: Oh, really?
Alex: Wow. How was it?
Randy: It was awesome.
Alex: Yeah, he’s the kind of performer who could pull that off.
Eric: We’re gonna have Alex rap the song now.
Alex: We will no,t but interesting side note, this guitar I play is the same kind of guitar that Ed Sheeran plays. He plays the three quarter sized guitar. So for those of you who are watching on YouTube, if we get this up on YouTube, this is Ed Sheeran’s guitar and I am by no means, Eric, equating myself to just head off your next comment with Ed Sheeran.
Eric: YouTubers, you are about to see the magic happen.
Alex: That’s right. Well that was the wrong chord. We’ll start again. [Singing]
Eric: Well done.
Alex: Is there Ed Sheeranish?
Eric: It’s just like Ed Sheeran.
Alex: You’re right, you’re right.
Eric: It’s just like being in Paris with Ed Sheeran. Right Randy?
Randy: I’ve got a tear in my eye.
Eric: So, we were hoping to talk to you about your recent JAMA IM article titled, “Effect of a Patient and Clinical Communication-Priming Intervention on Patient-Reported Goals-of-Care Discussions between Patients with Serious Illness and Clinicians”. That was a long title.
Randy: It was a long one.
Eric: So maybe before we talk about this particular article, maybe we can just take a step back and just think how did you get interested in this. Because this is really interesting. We’re using a priming intervention. Just tell us how you got interested in this, the topic of communication in the ICU.
Randy: Yeah. Well, this study, actually, we did …
Alex: It is the outpatient.
Eric: The outpatient, sorry. Sorry.
Randy: I got interested in this area a long time ago, in medical school actually, watching … I went to medical school in the late ‘80s and HIV had just hit Baltimore, and watched a lot of young men dying of a terminal disease at the time, and just seeing how incapable our healthcare system was at communicating with them. And then also simultaneously watching some amazing doctors do amazing jobs. So it was clearly that it can be done and yet most of the time it isn’t done well. And so I got very interested in trying to figure out how do we understand this better and how do we make it better.
Alex: Yeah. And this issue of priming-intervention, like patient activation, it’s a different take on how to promote advanced care planning, priming patients to talk with their physicians, and in this study, priming physicians to some extent as well. Well, where did the impetus for that component come from?
Randy: I also have come to believe that we really need to have a bilateral approach to this, that we need to prime both patients and prepare them for the conversation and prime physicians or nurse practitioners to do this as well. And I just felt like if we came at it at both angles and prepared both parties, we were more likely to get successful conversations happening.
Eric: So when we say prime, it’s not Amazon prime, right? What are we actually talking about when we say priming people?
Randy: I think this is a really important point because this priming, that’s preparing people, getting them to think ahead of time about … well, particularly patients, getting them to think ahead of time about what’s important to them and what their values are, and how they wanna talk about this; getting clinicians that information so that they’re also prepared. I think this priming intervention does not work, will not work, and can’t possibly work if clinicians don’t have basic skills.
Randy: So I believe that the SUPPORT trial, the famous support trial from the 1990s was a priming intervention that … My hypothesis is it failed because in the 1990s, this was not a basic skill that physicians had. I think, fortunately we’re not where we need to be but I think in 2018 the physicians and nurse practitioners have the basic skill and we can just prompt and prime, and help guide them, and we’ll get somewhere that we couldn’t have gotten in 1990.
Eric: And when you say basic skills, do you mean communication skills about goals of care?
Randy: Yes. Yeah. If they had no skill at all, then, this intervention, I believe, would fall flat and could even do harm. But if they have the basic skill, then helping them do a better job and prompting them to do it with this patient today, I believe, can be successful.
Eric: That’s terrific. So let’s talk about this particular study and talk about the setting in which this occurred. So could you talk a little bit about the population that you studied here?
Randy: Yeah. So we were interested in patients with a chronic, serious illness. We identified patients with a whole series of different chronic and serious illnesses, and we were targeting patients who, on average, had a median survival of about two years. And so, this is not advanced care planning with healthy patients, this is not planning for end of life care. It’s in the middle. It’s the beginning, early goals-of-care discussions for people in whom these decisions are not facing them today but might be soon. And we did this in the outpatient setting as well.
Eric: And how did you determine who fit into that two-year eligibility?
Randy: We had a series of different eligibility criteria for different diseases, oxygen dependency, OPD, metastatic cancer, interstitial lung disease with a total lung capacity of less than 50%. So for each class … for heart failure. For each disease we came up with criteria that we thought roughly put patients in that group.
Alex: And were these patients who had not previously had a goals-of-care discussion?
Randy: Well, that’s a good question and we did ask patients if they’d had a prior goals-of-care discussion, but we did not exclude them if they had. My recollection is about 40% told us yes.
Eric: And these are outpatients, and it was at multiple different institutions, is that right?
Randy: Yeah, it was throughout the Seattle area. So it included UW Medicine that includes two academic hospitals and two community hospitals, then we also went to a large community hospital system here as well.
Alex: And then you recruited outpatient docs to participate in this study?
Alex: Was that difficult or were they enthusiastic? What was that like?
Randy: Well, it wasn’t easy, I would say. We had a participation rate for the outpatient docs or nurse practitioners, included some nurse practitioners as well. Our participation rate was 27% which, I think, tells you right there that most doctors in their outpatient clinics are not signing up to do extra work. I’d like to believe that it would be much higher if it weren’t a study. If we were actually implementing this in practice, I think, we would probably get higher participation rates. I think part of the reason it was low is because we’re saying, “Hey, would you be willing to participate in this study? It’s optional to you and the good news, you get to fill out some surveys.”
Alex: Right, right, right. In that light, 27% actually may seem pretty good.
Eric: Yeah. That’s not bad.
Alex: Right, okay. So let’s talk about, the study design was a randomized trial.
Alex: Can you talk about the intervention and the control conditions?
Randy: Yes. So, for the intervention, what we did is we surveyed patients ahead of time to ask them questions about whether they had had these discussions, whether they wanted these discussions, what they saw as the main barrier to this discussion with their doctor and what their current values were and preferences were. And then we collated all that information into this one page, what we call a jumpstart tips form, which had two different forms.
One form went to the patient and said, “You have a doctor’s appointment coming up. We are encouraging you to have a conversation about goals-of-care. It’s easy. Here’s how you might start it, and here’s what you told us on the survey about your goals and preferences.” And then for the doctor, or nurse practitioner, a similar form but it said, “Your patient’s participating in this study. Either they are interested in having a goals-of-care discussion or they’re not, and here are their main barriers and ways that you might want to still address that, even though they told us they weren’t interested.”
Alex: Could you give one example of what kind of barrier was common and how might the clinicians overcome that. I understand you may have integrated with vital talk here.
Randy: Yeah, we did. Tony Bach was one of the investigators on this study and helped us a lot with thinking about how to address these barriers. So, for example, one of the barriers was, “I’d rather focus on staying alive than talk about end-of-life care.” Fairly common barrier and the suggestion was something along the lines of, “This is a difficult topic for this patient. Oftentimes patients who don’t want to talk about this, it’s better to approach it indirectly and talk about patients with a disease like you rather than you.” That’s just one example.
Eric: And remind me how did the clinician get that feedback again, how to address this?
Randy: They got a one page … It was a piece of paper they got by email one or two days before the clinic appointment and then a paper copy showed up in their box in the clinic on that day.
Eric: And the suggestion was also in there and how to deal with …?
Randy: Yeah. It’s a one page. We have examples on the website. It’s a one page document that says … provides, maybe, a little too much information, frankly, that’s some of the feedback we got but it provides all this information on one piece of paper.
Eric: In addition to that, did the clinicians get any other goals-of-care training on how to have these discussions?
Randy: We didn’t control that, so many of them in our system have had some training and some will have not. We don’t have a systematic way, or at least we didn’t when this study was being done. We didn’t have a systematic way to train physicians in this that was in place. So that was … that would be very hit and miss.
Alex: And what were in the control condition in this case?
Randy: So the control condition, I think, is important to understand because it was what I like to think of as, sort of, usual care-plus or enhanced usual care in the sense that the control patients also filled out the same surveys that asked them about having these conversations, barriers to these conversations, what’s important to them. And I believe that those surveys actually prompted some discussions. And so, if you look at our control group, 31% of the control patients had a discussion about goals-of-care on a routine clinic visit, which I believe is much higher than in routine practice. I think it’s much, it’s probably less than 10%. So I do think there was some prompting that happened by virtue of filling out the survey.
Alex: It was like a mini jumpstart.
Randy: Yeah, it’s like part of a jumpstart.
Alex: I don’t what that would be.
Eric: It also reminds me a little bit of Rebecca Sudore’s work with Prepare for Your Care.
Eric: Kind of, flipping the classroom a little bit, getting people to think about this before they talk to their physicians. Rebecca’s work also includes a lot of videos, working through five different steps and patient friendly formats to get people to start thinking about their goals-of-care and create an action plan. It sounds like there’s components of that here. Is that right?
Randy: Yes, I would say that that’s right. I would say that the patient-facing part and patient-prompting part is similar. Although in our intervention I would say it’s much less detailed. That’s really the focus of her intervention, is patient education and engagement and for us, I think, it was a smaller part and a shorter process for the patients.
Alex: And the focus of this intervention was more about priming the patient and the doctor to have a high quality conversation, high quality communication in the office, doctor or nurse practitioner, the clinician. Whereas Rebecca work is much more focused on doing pretty much everything in the outpatient setting and, “Yeah, it’d be great if you brought this to your doctor as well.”
Randy: Yeah, although her work has also prompted these discussions. So she showed it, in one of her recent randomized trials, that having the patients do Prepare For Your Care and then asking them to bring them into the clinician did actually … they did that and it prompted discussion. So I do think there’s some overlap.
Eric: So, that’s … Now maybe talk a little bit more about what you actually found in this study, doing priming.
Randy: Okay. Yeah. So our primary outcome was whether, in fact, we could prompt these discussions and we looked at that … The primary way we looked at that was to ask the patients. We also looked at what the clinicians documented in the EHR and, in addition, we ask the clinicians on surveys whether they talked about it. Our primary outcome, we did see an increase in patient report of these discussions, 31% in the control group and 74% in the intervention group. We also saw an increase in clinician documentation in the EHR from 17 % in the control group to 62% in the intervention.
Eric: Wow, those are big differences there.
Alex: Yeah. It’s hard to say that there’s nothing going on because there’s definitely something going on there.
Randy: Yeah, yeah. I was very pleased with that. I thought that was really terrific. And then we also asked patients to rate the conversation that they had with their clinician using a survey instrument called The Quality of Communication and the patient’s rating of the quality of the discussion also increased significantly.
Alex: And you had another outcome here that is, sort of, the holy grail of palliative care, which is goal concordant care. Can you talk about … and then you have the issue going … and I’m sure you and many others who may be listening are aware of this, how do you measure that? And you chose to measure it a specific way based on SUPPORT. Could you talk about that selection?
Randy: Yeah, so I think you’re right that it is the holy grail of how we measure goal concordant care. And I would be the first to say, first in a long line of people to say, that there’s no perfect way to measure this and that there are real conceptual challenges to measuring whether care delivered is goal concordant for lots of reasons, an important one being that patient’s goals can change. We chose to measure it with this question that was first developed for the SUPPORT trial that we talked about earlier done in the 1990s, that asks patients to pick between two choices.
What is clearly, in a sense, a false dichotomy but it asks them to choose, if you had to choose right now, would you prefer a course of care focused on comfort, even if it means not living as long, or a course of care living as long as possible, even if it meant not being as comfortable or not being comfortable. And it’s a false dichotomy because, of course, everybody wants both, right? Everybody wants to live as long as possible, as comfortably as possible but it’s an approach that I personally like. Most of my colleagues hate it, most palliative care doctors hate it, and most palliative care researchers hate it.
Randy: I like it, and I like it because it’s not a real choice. It’s helping you understand somebody’s priorities when faced with tradeoffs. And so I think, as a research tool, I think it’s actually, potentially, quite helpful. And so we asked them first what would you prefer and then we say, given the same two choices, how is your care focus now? And we defined goal concordant care as care in which they say, “This is what I prefer and my care is focused in the same way.”
Eric: Well, it’s interesting too because I actually use that a fair amount in my clinical work is, and I think I got this from a fellow who got this from Rebecca Sudore, is using a spectrum. Like on one hand we have patients that we care for who would want to do everything they can to try to prolong their life, even if it causes some discomfort. On the other hand, we have patients whose primary goal is just to be comfortable, even if it means they may not live as long. And then in the middle we have most patients where they fall somewhere in the middle. Where do you fall? It sounds like it’s a similar thing here; we’re just taking out the middle. If you had to make a choice right now, where would that be?
Randy: Right. And I think from a research and measurement perspective, taking out the middle is really important because the vast majority of the population, of course, is in the middle.
Alex: Yeah, right.
Randy: Whether you’re a 3.8 or a 4.2 on a five point scale is pretty meaningless from my perspective but when you do this, sort of, forced choice exercise most people will put their money down on one or the other. There are some people who won’t, who say I don’t know and we chose, from a research perspective, to define them as not proceeding goal concordant care. But most people in the study chose one or the other and, frankly, the majority chose comfort over longevity.
Alex: And it’s in contrast to something Steve Pantilat said on this podcast when we asked how do you define palliative care? And he says, “Clinically, I say to patients, my goal is to make you … is to help you live as well as you can for as long as you can,” which is, sort of, the synthesis of those two goals together. I wanted to ask you about, just sticking with the goal concordant care, if you could talk about the outcomes there in this study.
Randy: Yeah. So we looked at it overall in all patients and we did see an increase in the proportion who reported goal concordant care at three months after the intervention, the target visit where the intervention was delivered. But it wasn’t statistically significant, the p-value was 0.08. And then we looked at an a priori chosen subgroup of patients who goals were stable, between three months and the last time we measured it. And we chose that sub group because, this is what we talked about before, that if patients’ goals are changing, it’s pretty hard to align their care with their goals.
And so we said, well, if their goals were stable, did we see a significant improvement? And there we did. We saw a significant increase in goal concordant care amongst that subgroup with stable goals.
Alex: And so this is really an exciting encouraging study. I guess my next question is how much does it actually mean? Because, ultimately, what we care about is improving the quality of these conversations so that we can improve patient experience in the setting of serious illness. So we have some signals from the patients that maybe there was more goal concordant care and we certainly have improvement in communication, improvement at documentation. What are the next steps here, in the outpatient setting, and how far … what would you ideally measure if you could?
Randy: Well, I think you’re absolutely right that that’s the key question is, does this change care or does this change the patient’s experience of care? And we don’t know, based on this study, whether it changed care. We had almost 250 patients and yet only about 60 died in the six months that we followed them, and so we really weren’t powered to be able to see whether it changed the care they got at the end of life. That’s what I’d like to know. I’d like to know, did it change the care they got and did it change their experience of their care at the end of life? And you’d need a much bigger study. I also think, I see this as sort of, a proof of concept study. I also believe that it’s, probably, naive to think that doing this once over the course of a patient’s illness is gonna have dramatic effects on the care they receive. I think we know from our experiences, as palliative care providers, that these conversations are not things that you generally can have once and be done with, but rather something that needs to be continually updated.
Eric: But let’s say you sold me Randy. Let’s say I’m really, I’m down for priming. Is this ready for primetime? Should I be using this? How can I use this?
Randy: Yeah. So if the goal, if you say to me, “I’m in charge of a healthcare system, the healthcare system will do whatever I tell it to and I believe that having these conversations is important and I want these conversations to be happening in my healthcare system.” Then I say yes, it’s ready.
Eric: Because I work at the VA, and nationally at the VA we have a new policy around life sustaining treatments where any high risk patient, pretty much everybody who would qualify in this study would qualify for our policy, should have goals-of-care discussions. What would that rollout look like if we wanna do that VA-wide?
Randy: So to roll this out VA-wide you would need to have in place, which you do, a system to send people surveys and to collect the results and then it can automate these jumpstart forms. And then you’d have to have a system in place to get the jumpstart forms to the patients and the clinicians before a target visit.
Eric: And let’s say I’m interested in these jumpstart forms. Are they available? Can I actually get them somewhere?
Randy: Yeah, there’s examples of both how we created them, the algorithm that we used to go from the survey to the jumpstart form on the JAMA internal medicine website. And then there’s also a two examples, a clinician example and a patient example.
Alex: And what proportion of patients actually did everything, completed all the surveys? In other words, do you have a sense of how burdensome this was and what the sorta uptake was on the patient side?
Randy: Yeah. So that’s a good question and that would be a big issue if you did what I just said as a VA system and said, we’re going to send these out, we’re going to get them back, we’re going to put them in. Within this research study, with funding from PCORI to do this study and staff to send these out and bug people to send them back, we had 59% of eligible patients agree to participate. And then we had survey results, depending on when, somewhere between 70% and 80% of the patients.
Randy: So with that staff in place, you could reach somewhere around 50% of eligible patients. Reach meaning, they fill out the survey, they send it back and you get them the jumpstart form. But I think that the key there is with that staff and what I just said to you about what you would need to roll this out in the VA didn’t include that level of staff.
Eric: It sounds like, you mentioned before, there’s some lessons learned that you have, like maybe the forms are a little bit too much or too detailed.
Randy: Yeah, we did qualitative interviews with patients, family members and clinicians to get feedback on the intervention and that was a common piece of feedback. It was that it was too much, there was too much information there. So in our next iteration of this, I do want to simplify the forms and I also think that the tips that we provide, I think they were good, but I think they could be improved. And so if I were gonna roll this out in a healthcare system, I’d bring together palliative care experts and others and say hey, look at these tips and tell me how to make them better.
Alex: I saw that you, congratulations, you got an NPCRC pilot award and you’re going to be studying this same intervention, my understanding is, in an inpatient setting.
Randy: That right.
Alex: I wonder if you could tell us more about that next step, speaking of next steps, and why you decided to move from the outpatient setting to the inpatient setting.
Randy: Yeah, so for me in this intervention, I think there are really two next steps. One, as you’ve said, is to try this on the inpatient side and we’re very appreciative to NPCRC for handing this pilot. I think it really needs to be piloted because I think we don’t know how this is gonna work on the inpatient side. The inpatient setting is very … everything is time compressed, everybody’s very busy. I do know that, from our preliminary data, that there are a lot of patients who are admitted to the hospital who should have a goals-of-care conversation and don’t.
So I believe the opportunity is there to prompt these conversations in the inpatient but whether we can successfully do that in a setting where, at least in our city, everybody’s focused on getting the patient discharged as soon as possible. And so I think that’ll be a challenge and I’m excited to try it on the inpatient side.
Alex: But sometimes you find, surprisingly, that having these conversations, especially early in the course of the hospitalization, may actually lead to shorter lengths of stay.
Randy: Yeah, I think that’s right and it may lead to palliative care consults and hospice referrals and I think that’s right, and that would be the goal to look to see if we can do that. The other thing that I really want to do, since you asked about next steps, and have written up a PCORI, another PCORI grant to do that, is to compare. So this intervention requires that we get ahold of patients and survey them and then use that information in the jumpstart.
What if we just designed a more generic jumpstart that didn’t have to go through that whole process and compared the patients’ specific jumpstart to a more generic jumpstart? And that’s what we want to do in a comparative effectiveness trial because that would be a lot easier to roll out in a healthcare system, if it was as effective. But I don’t know, I don’t know whether it will be as effective.
Eric: My last question is, and we’re talking about next steps, and I know that the next step that you really want is Alex to do a little bit more Ed Sheeran.
Randy: That’s true.
Eric: But is there anything else that you’d like to talk about while we have you online?
Randy: I guess the other thing that I would say is that, I think that there are a lot of exciting interventions out there, like Rebecca Sudore’s Prepare For Your Care that you mentioned, like the serious illness checklist that the group from Harvard is doing, like the Vital Talk teaching. What I would think would be fabulous is that, I think that all of those things could be used together. I don’t think it’s an either or, I don’t think they’re in competition but they all take, somewhat different approaches. How do we take the best of each and create a system that we’re really prompting, guiding and helping these conversations to be the best they can be. That’s my dream for 2020.
Eric: I think that would be great because as a provider, sometimes you feel a little bit overwhelmed when there are too many choices of what should I do? Should I do Rebecca’s PREPARE? Should I do priming? Should I do..
Alex: Yeah, the Vital Decisions, a video decision aid, Angelo Volandez who was on this podcast.
Eric: And nevermind, I’ll just go home. There’s too many questions, I can’t take it.
Alex: Paralysis of choice?
Eric: Yeah, we’ll focus on diabetes targets instead.
Alex: Well, that’s great. Thank you so much Randy.
Randy: Yeah, I enjoyed it. Thanks for having me.
Eric: Well, how about before we end Alex, do you want to give us a little bit more Ed Sheeran?
Alex: Sounds good. [Singing]
Eric: Great, I couldn’t tell if you were done with that. Thank you Randy again for joining us and all of our listeners for joining us this week, we look forward to our next podcast next week too. If you have a moment, please take a little step and just rate us on your favorite podcasting software, whether it be iTunes or Stitcher or anything else. And with that, we’ll end. Thank you.
Alex: Thanks Randy. Bye.
Randy: Thanks, Bye.